Tanja Farrokh-Eslamlou (Presenter)
Imperial College London
Bio: I graduated from King’s College London in July 2016, having obtained a BSc (Hons) in Biomedical Science. During my time here, I chose to study optional modules which included the neurological and physiological changes that occur in the event of metabolic dysregulation as a result of obesity, as well as its resultant malignancies, including cancer. In my final year project, I decided to team the physiology of obesity together with my other passion: the phenomenon of reproduction and pregnancy. Following on from this, I began my postgraduate studies in October 2016 at Imperial College London, where I am currently studying towards an MSc in Reproductive and Developmental Biology. My ongoing project utilises the exciting new DESI-MSI technology to combine obesity, metabolic dysregulation and endometrial carcinogenesis, paving the way towards a more thorough understanding of tumorigenesis
Authorship: T Farrokh-Eslamlou (1), O Raglan (1,2), L Doria (3), R F Soares (3), F Rosini (3), J McKenzie (3), M Kyrgiou (1,2), Z Takats (3)
(1) Institute of Reproductive and Developmental Biology, Department of Surgery & Cancer, Imperial College, London, UK, (2) Imperial Healthcare NHS Trust, London, UK, (3) Computational and Systems Medicine, Department of Surgery and Cancer, Imperial College, London, UK
Endometrial cancer (cancer of the womb lining) is the most common gynaecological cancer diagnosed in Europe. Strongly associated with obesity, the number of new cases of endometrial cancer being diagnosed is rising. DESI-MSI is an ambient ionisation technique that allows direct correlation between biochemical changes and histological features within a tissue. Detection of tissue-specific lipid ion patterns enables a novel method for tumour biology investigation by providing unique, biochemical information. This pilot study demonstrates the successful application of DESI-MSI as a tool for accurate characterisation of endometrial tissues.
Endometrial cancer, a tumour originating in the lining of the uterus (womb), is the most common gynaecological tumour in developed countries, and its prevalence is rising. Although endometrial cancer is commonly assumed to be a cancer of the postmenopausal period, 14% of cases are diagnosed in premenopausal women, 5% of whom are younger than 40 years.
Endometrial cancer (EC) diagnosis, treatment and prognosis currently relies upon peri- or post-operative analysis of solid tissue biopsies and a multidisciplinary team approach. This includes surgical excision of the biopsy and surrounding tissue area, and preparation of tissue sections for Consultant Histopathologist identification of cancer subtype and grade. The manual and multidisciplinary approach involving numerous departments for one patient sample is extremely time consuming. Additionally, this gold standard method is subjective, and therefore may be subject to human-error and misjudgment. There is a need for novel time and resource-effective tissue identification methods able to robustly identify benign versus endometrial tumour grade subtypes.
All biological membranes include lipids in their biochemical structures, which are fundamental from a structural and functional aspect, but they are also an important class of metabolites. These metabolites have been reported to become altered within tumour biology, suggesting a use of these lipid metabolites as biochemical markers of cancer progression, providing a crucial, deeper understanding of cancer biology and therefore, potential improvements to patient treatment and prognosis.
DESI-MSI can generate spatially resolved metabolic profiles, using individual mass spectra collected from the tissue samples. These individual mass spectra can be used to investigate the distribution and relative abundance of specific mass ions within tissues, providing a potential tool to aid the current methods used by histopathologists. The strength of DESI-MSI is in its ability to process samples in ambient conditions, with minimal sample preparation and robust detection of tissue-specific lipid ion patterns. This pilot study aimed to use desorption electrospray ionisation mass spectrometry imaging (DESI-MSI) to characterise and differentiate between distinct tissue types within benign and high grade endometrial tumour tissues, based on the unique lipidomic profile of the individual tissue types.
Fresh frozen sections of benign or endometrial cancer tissues were cryosectioned (10 µm thickness) and mounted onto a 2D-linear moving stage. High Definition Imaging 1.4 software (Waters Corporation) was used to define the tissue area to be analysed by DESI-MSI. Mass spectra were collected at specific x and y coordinates at a pixel resolution of 100 µm. The experiment was carried out in negative/positive ion modes (mass range 50-1,500).
An imaging toolbox developed in house was used for raw imaging data processing and areas of interest annotated with the haematoxylin and eosin optical image. Unsupervised and supervised statistical analysis were performed for tissue classification.
DESI-Imaging was performed on a pilot cohort of 15 samples, including 8 benign endometrial tissue samples and 7 high grade endometrial cancer samples (grade 3).
Using the in house imaging toolbox software, we compared and directly overlay the ion images generated using DESI to the haematoxylin and eosin (H&E) stained images, annotated by a histopathologist. Different ions produced different images of the tissue section which could be directly correlated with the spatial distributions across the sample surface. From the spectra collected, we were able to identify specific fatty acids and glycerophospholipids. Tissue type differentiations where the regions of interest on the normal (benign) endometrial tissue and the endometrial tumour could be identified were then separated by principal component analysis (PCA, unsupervised analysis) using the imaging toolbox. This showed that benign myometrium and endometrial cancer grade III tissue clusters are clearly separated and distinct from one another. Further supervised analysis using maximum margin criteria (MMC) was carried out, showing that discrimination of benign myometrium versus tumour (grade 3) was possible, with a leave-one-patient-out cross validation accuracy of 87.5% for grade 3 tumour samples and 85.7% for benign myometrium.
Putative identification of m/z values based on accurate mass was carried out using Metlin (http://metlin.scripps.edu) and Lipidmaps (http://www.lipidmaps.org). It was identified that PA (36:1) was more abundant in benign endometrial tissues, comprised of myometrium. PG (O-36:2), PE (O-36:0) and PI (O-36:0) held a higher abundance in high grade aggressive endometrial cancer (grade 3) tissues.
Conclusions & Discussion
This pilot study provides data to suggest that DESI-MSI has the potential to provide reliable, accurate differentiation of endometrial tissue types by identifying a unique lipid signature among benign endometrium and malignant high grade endometrial tumour.
References & Acknowledgements:
IP Royalty: no
|Planning to mention or discuss specific products or technology of the company(ies) listed above:||